| Title |
Variable Step-Size P-norm-like Affine Projection Algorithm |
| Authors |
유진우(JinWoo Yoo) ; 박범용(Bum Yong Park) ; 신재욱(JaeWook Shin) |
| DOI |
https://doi.org/10.5370/KIEE.2020.69.2.344 |
| Keywords |
variable step-size; affine projection algorithm; impulsive noise; P-norm |
| Abstract |
This letter presents a p-norm-like affine projection algorithm (APPA) and its variable step-size algorithm to improve the performance in high probability impulsive noise. The APPA is derived by minimizing the p-norm-like of an error vector that causes reducing error vector, when impulsive noise occurs. Therefore, the proposed algorithm has low steady-state estimation error in impulsive noise environment. In addition, its step-size algorithm is proposed by minimizing the mean-square deviation of the APPA to improve the performance in terms of the convergence speed and steady-state errors. The proposed algorithm is tested in the channel identification scenario including high probability impulsive noise. Simulation results show that the proposed algorithm has faster convergence rate and lower channel estimation errors than the variable step-size affine projection sign algorithm in impulsive noise environment. |